Reliability assessment of optoelectronic and photonic devices in - - PowerPoint PPT Presentation
Reliability assessment of optoelectronic and photonic devices in severe environments: architecture and applications of the OpERaS consortium OpERaS : Op to- E lectronic R eliability a pplied to S evere environments L. BECHOU, P. SPEZZIGU
1 1
Laboratoire IMS University of Bordeaux 1 CNRS UMR 5218
phone : +33 5 4000 2767 laurent.bechou@ims-bordeaux.fr piero.spezzigu@ims-bordeaux.fr
2 2
RADIATIONS + HIGH THERMAL CYCLING : -160° C/+150°C + VACUUM (10-9 torr) + VIBRATIONS (± 10g) + ELECTROSTATIC DISCHARGE
3 3
Solid-State Lasers: LIDAR, metrology, interferometer, atomic clocks… Detectors (X -> IR) and fiber sensors: T, P, mechanical stress Optical links and interconnects: Intra-&Inter-satellite, Inter-chip Photonic Signal Processing, Non-linear optics: optical data storage, µwave generation, MOEMS…
Astronomy / Planetary Exploration, Fundamental Science Earth Observation, Remote Sensing Telecom - Navigation Space Transportation ISS
Source : E. Armandillo, Space Optoelectronic Day, Cork, 2006 Source : Courtesy of Lumics GmbH, Berlin
4 4
– tests according to MIL, Telcordia or electronic-related ESCC standards
– Cost reduction
– Tight time schedule
– Small procurement volume (i.e. few tens of devices)
– Specific environments (i.e. radiation, vacuum, high temperature range) – Few amount of reliability data (especially true for custom devices or low volume productions)
5 5
– Standard tests are generally not well-fitted for the devices – Due to the small number of devices the objectives of the evaluation/qualification phases are not obvious (Could we rely on these tests to guarantee the reliability of the device during the space mission ?
How could we prove it ?)
– Difficulty to find companies that could take in charge evaluation/qualification tests (The manufacturers are not always interested due to the low procurement volume) – Many anomalies during evaluation or qualification phases – Difficulties to identify physical root causes and prediction of operating lifetime is traditionally extracted from empirical laws using classical methodologies (t50% , MTTF…)
6 6
Failure A (often unknown root cause) Device to qualify Standard evaluation
Qualification Best-fitted evaluation
Strong over-cost due to :
Failure B (often unknown root cause)
THE "VICIOUS" CIRCLE FOR COTs DEVICES RELIABILITY !
7 7
Round table "Main issues to be solved" - 12 may 2006 Workshop "Laser Diodes for Space Applications" (CNES Toulouse-France) Round table "Main issues to be solved" - 12 may 2006 Workshop "Laser Diodes for Space Applications" (CNES Toulouse-France)
8 8
9 9
"Safe" qualification area assessment
"Safe" qualification area assessment
Technological weakness points identification
(semiconductor, assembling processes)
Technological weakness points identification
(semiconductor, assembling processes)
10 10
Final client Space charge manufacturer Space instrument manufacturer Optoelectronic components fabrication Operator Satellite or space vehicles On-board equipments Components
Reliability Assessment, Tests, Qualifications
11 11
Steering Committee
CLIENT, AdvEOTec, IMS, TISS, Manufacturer
AdvEOTec Partner 2 Partner 1 Partner i
CLIENT
Optoelectronic end-user
OpERaS Project (n°i)
Tests A Tests B Tests C Tests Z Lab A IMS, TISS Analysis 1 Analysis 2 Analysis 3 Analysis n Lab B Lab N
Manufacturer
Cross-correlation in metrology and measurements Data capitalization in a specific data base (respect to NDA)
Evaluation & qualification tests management Evaluation & qualification tests management Expertise and research management, Technological analysis and failure mechanisms modelling Expertise and research management, Technological analysis and failure mechanisms modelling (www.eurelnet.org)
12 12
GALILEO Project : Study of 850nm silicon photodiodes degradation under radiation effects : Gamma (ionizing dose), Protons (displacement dose)
S1337 BQ/BR Hamamatsu (detection, dosimetry)
P+- 1µm SiO2 Incident light N+-1 µm
ph
N - 320 µm
13 13
Client : CNES Selected component: PIN 850nm Si photodiode (COTs) Application : Galileo (atomic clock) Client : CNES Selected component: PIN 850nm Si photodiode (COTs) Application : Galileo (atomic clock) Initial characterizations (I-V, S-λ, Linearity, NEP) Pre-evaluation phase (radiation effects) "Light" Pre-evaluation phase Bibliography Failure mechanisms Operating lifetime Qualification phase Modelling Initial characterizations (I-V, S-λ, Linearity, NEP)
(extraction of dopants concentration)
14 14
0,01 0,1 1 10 100 1000 0,01 0,1 1 10 Thickness (mm Al) lifetime (years)
Lifetime (Years) Shield thickness (mm Al)
Results published at ESREF 2008 (Maastricht) and in JAP,
From physical failure mechanisms… From physical failure mechanisms… … to reliability … to reliability
0,1 0,2 0,3 0,4 0,5 400 500 600 700 800 900 1000 1100 Wavelength (nm) Responsivity (A / W) initial measurement Dose 0,1 0,2 0,3 0,4 0,5 400 500 600 700 800 900 1000 1100 Wavelength (nm) Responsivity (A / W) initial measurement Dose
Under protons => ΔR > 25% @ 852nm Responsivity (A/W) Wavelength (nm)
Before radiations
1,0E-14 1,0E-13 1,0E-12 1,0E-11 1,0E-10 1,0E-09 1,0E-08 1,0E-07 0E+00 1E+08 2E+08 3E+08 4E+08 Dose = * NIEL (MeV.g-1) Idark @ -0,1 V (A) @ 300 K @ 283 K @ 253 K Model 1.108 2.108 3.108 4.108 1.10-8 1.10-9 1.10-10 1.10-11 1.10-12 1.10-13 1.10-14 1.10-7
Idark @ -0,1V (A) Dose = Φ x NIEL(E) in MeV.g-1
] D .[ k 1 1 . D L
d d , g init irrad
+ τ = τ τ =
⎟ ⎟ ⎠ ⎞ ⎜ ⎜ ⎝ ⎛ − ⎟ ⎠ ⎞ ⎜ ⎝ ⎛ ⋅ τ ⋅ ≈ 1 kT qV exp D D N qn J
p p p D 2 i ) d ( dark
( )
1 kT 2 qV exp 1 kT qV exp V V qN 2 2 qn J
s D g i ) g ( dark
+ ⎟ ⎠ ⎞ ⎜ ⎝ ⎛ − ⎟ ⎠ ⎞ ⎜ ⎝ ⎛ ⋅ − ε ⋅ τ ≈
15 15
CNES/AdvEOTec-IMS
Reliability of Si phototransistors (2007-2010) : CNES/IMS-Univ. Cagliari (1 PhD defended) Evaluation of 0.98µm pump Laser diode for space applications (2008-mid 2010):
CNES/AdvEOTec-IMS-TISS/3SPhotonics (1 PhD supported)
Evaluation of 1.55µm DFB Laser diode for space applications (2009-mid 2011)
CNES-ASTRIUM-Thales Alenia Space/AdvEOTec/3SPhotonics
Reliability assessment of commercial optocouplers (2009-2010)
CNES/AdvEOTec-IMS/Micropac (USA)
Failure analysis of 1.55µm Laser diodes under EOS/ESD tests (2009-mid 2011)
CNES/AdvEOTec-IMS-TISS/3SPhotonics
16 16
New generation of spectral-resolved cathodoluminescence imaging system for
New architecture of rad-hard optocoupler using VCSEL technologies (2011-
IMS-IES/3SPhotonics
Embedded passive devices in multilayer PCBs for high-frequency applications:
(IMS/AdvEOTec/Polyrise - EDA Projects and Programmes – ITP SIMCLAIRS in evaluation)
17 17
Cryogenic bench L-λ, P-I, Far-field (chip on submount) Δλ, RIN, Chirp, frequency modulation measurement (Fiber Laser diode) CW & pulse I-V, P-I tr, tf, C-V Photodetectors Photodetectors Emitters Emitters
Spectral sensitivity, CTR, linearity and Gummel-plot
18 18
19 19
20 20
AFM imaging on COD (CRPP, Bx) CL imaging (THALES ISS) TEM imaging (THALES ISS)
Dislocations
21 21
22 22
Issues in phototransistors degradations in space environment physical modeling approach:
23 23
Design of Experiments is a structured, organized method for determining the relationship between factors affecting a device characteristic and the characteristic itself
inputs
Responses
temperature. bias radiations
Factors
24 24
TID and DDD: the experimental damage factors The study domain boundaries depend on test facilities capabilities: MeV/g 10 MeV/g 10
9 6
≤ ≤ DDD
TID and DDD deposition with γ‐rays and protons irradiation → TID is also deposited by proton irradiation The Domain in which the couple (TID,DDD) could be deposited with only proton irradiation is defined by (using a logarithmic scale):
11
(1) log( ) log( (184)) log(1.6 10 ) log( (184)) log( ) TID LET NIEL DDD
−
≥ + ⋅ − +
11
(2) log( ) log( (30)) log(1.6 10 ) log( (30)) log( ) TID LET NIEL DDD
−
≤ + ⋅ − +
Proton energies: 184MeV and 30MeV krad 100 krad 1 . ≤ ≤ TID
25 25
GEO/MEO/LEO mission profiles (OMERE)
by a DoE software and based on a “optimality criterion”
6 7 8 9
0.0 0.5 1.0 1.5 2.0 2.5
LEO - 800km/800km - 98° MEO -1000km/26768km - 63.4° GEO - 35870km - 0° LEO 5yrs LEO 25yrs MEO 5yrs MEO 25yrs GEO 5yrs GEO 25yrs DoE points
log(TID[krad]) log(DDD[MeV/g])
Study Domain Increase of Al shield thickness
(eq.2) (eq.1)
26 26
Pre-defined Response function :
where x=log(DDD) and y=log(TID). R(x,y) : response of interest (i.e. photocurrent, darkness current, or Sp. Responsivity) B=Z0,A,B,C,D,F : the unknown coefficients of the polynomial.
→ → 9 Equations (for 9 experiments) and 6 unknown system: 9 Equations (for 9 experiments) and 6 unknown system:
2 2
exp 1
T T
−
27 27
Irradiation sessions: “OFF” and “ON” Responses:
Photocurrent Darkness current Spectral Responsivity
Methodology validation: Beam degrader target mission simulation
800km – LEO – inclination of 98° – 7mm‐thick spherical Al
50 100 150 200 10
8
2x10
8
3x10
8
4x10
8
5x10
8
calculated with OMERE simulated using the beam degrader
Differential fluence (MeV
Proton energy (MeV)
The target DDD and TID are: 1.8∙108MeV/g and 9.32krad(Si)
corresponding logarithm values: 8.2553 (log(DDD)) and 0.9694 (log(TID))
NB: Dose Rate Sensitivity evaluation: not relevant
28 28
6 7 8 9
1 2 0,2 0,4 0,6 0,8 1,0
"OFF" experiences DoE experimental data Beam degrader data
I
P H
/ I
P H
log(DDD) l
( T I D )
Normalized photocurrent: IPH/IPH0
/ (8.2553,0.9694) 0.531
PH PH
I I =
Predicted value: Predicted value:
6 7 8 9
1 2 0,2 0,4 0,6 0,8 1,0
"ON" Experiences DoE Experimental data Beam degrader data IPH/IPH0 log(DDD) l
( T I D )
/ (8.2553, 0.9694) 0.5219
PH PH
I I =
Predicted values in agreement with Beam Degrader simulation
29 29
Optocoupler CTR degradation
Predicted value: ‐2%
6 7 8 9
1 2
0%
"O N" Experiences DoE Experimental data Beam degrader data
Δ
CTR log(DDD) log(TID)
6 7 8 9
1 2
0%
"ON" Experiences DoE Experimental data Beam degrader data
Δ
CTR log(DDD) log(TID)
Type 1 (non rad‐hard) Type 2 (rad‐hard)
Predicted value: ‐14%
Predicted values in agreement with Beam Degrader simulation
30 30
Reducing number of experiments
Predicted values in agreement with Beam Degrader simulation
Study Domain
(1) GEO 5Y ‐ 6.75m m (2) GEO 20Y ‐ 6m m (3) Spot 18Y ‐ 7m m (4) GEO 5Y ‐ 40m m ‐1.5 ‐1 ‐0.5 0.5 1 1.5 2 2.5 5.5 6 6.5 7 7.5 8 8.5 9 9.5
log(DDD)
log(TID)
1 2 3 4
6 7 8 9
0% 20%
1 2 DoE Data Beam Degr. Data
Δphotocurrent
log(TID) log(DDD)
31 31
From OMERE database: ionizing and displacement doses for different mission profiles
Perigee Apogee Inclination Period # of orbits LEO (Low Earth Orbit) 800km 800km 98° 6043s 100 MEO (Medium Earth Orbit) 1000km 26768km 63.4° 28689s 300 GEO (Geostationary Earth Orbit) 35870km 35870km 0° 86400s 1
6 7 8 9
0.0 0.5 1.0 1.5 2.0 2.5
LEO
/800km
5yrs 10yrs 15yrs 20yrs 25yrs
lo g (T ID ) log(DDD)
6 7 8 9
0.0 0.5 1.0 1.5 2.0 2.5
M EO
5yrs 10yrs 15yrs 20yrs 25yrs
lo g (T ID ) log(DDD)
6 7 8 9
0.0 0.5 1.0 1.5 2.0 2.5
G EO
5yrs 10yrs 15yrs 20yrs 25yrs
lo g (T ID ) log(DDD)
Extracted photocurrent polynomial from “ON” irradiations results:
2 2
( , ) 0.636 0.385 0.059 0.069 0.052 0.692
PH PH
I x y x y x y x y I = ⋅ − ⋅ − ⋅ − ⋅ + ⋅ ⋅ −
32 32
10 20 30
0%
LEO - 800km/800km - 98° Photocurrent Degradation Mission Duration (years) 3mm 5mm 7mm 10mm 15mm 20mm 30mm 50mm 100mm
20 40 60 80 100
0%
LEO - 800km/800km - 98° Photocurrent degradation Al Shield Thickness (mm) 2yrs 5yrs 10yrs 15yrs 20yrs 25yrs
Example: LEO mission profile Photocurrent degradation vs. mission duration (for different shield thicknesses) or vs. the shield thickness (for different durations) The same charts could be easily obtained for MEO or GEO mission profiles
33 33
Failure criterion: 40% photocurrent degradation
These charts provide a quick overview, for qualification purposes, of the amount of shielding needed for specific orbit and a fixed duration, for a fixed failure criterion.
5 10 15 20 25 30 10 20 30 40 50
LEO M EO G EO
A l S h ie ld th ickn e ss (m m ) M ission Duration (years)
34 34
Reliability data extraction From DoE data, OMERE mission profiles data, and considering a 40% of photocurrent degradation, using a lognormal distribution we obtain the cumulative function plots:
7 14 21 0.0 0.5 1.0
LEO F experimental F(t) LogNorm C um ulative de sity fun ctio n t f (years) 14.16
8 12 16 0.0 0.5 1.0
MEO F Experimental F(t) LogNorm Cumulative desity function tf (years) 11.90
20 24 28 0.0 0.5 1.0
23.68 GEO F Experimental F(t) LogNorm Cumulative desity function tf (years)
ln 1 1 ( ; , ) 2 2 2
f X f
t F t erf μ μ σ σ − ⎡ ⎤ = + ⎢ ⎥ ⎣ ⎦
lognormal distribution CdF: lognormal distribution PdF:
2 2
(ln ) 2
1
( ; , )
2
x X
f x e
x
μ σ
μ σ
σ π
− −
−
=
2
2 m
MTTF t e
σ
=
9
( ) ( ) 10 1 ( ) f t t FIT F t λ = −
devices mean time to failure devices failure rate
35 35
We have demonstrated that statistical Design of Experiments is a very useful tool to evaluate BSPA degradations in space environment
– It needs basic knowledge of device and device degradation physics – As device is described with the “black box” approach, it could be extended to other families of components – It provide a tests plan depending on test facilities capabilities – With only one tests session, it gives the BSPA degradation previsions relative to a wide range of possible space mission profiles – Could be optimized (reducing number of experiences)
36 36
AdvEOTec AdvEOTec – F. Rosala (CEO)
Expertise and experience in Expertise and experience in
+33 (0) 1 6086 4361 francoisrosala@adveotec.com www.adveotec.com
IMS IMS – L. Bechou (Full Prof.)
Laboratory Laboratory with with expertise and expertise and experience in optoelectronic reliability experience in optoelectronic reliability for space and telecom applications* for space and telecom applications* +33 (0) 5 4000 2767 laurent.bechou@ims-bordeaux.fr www.ims-bordeaux.fr
THALES ISS/CEL THALES ISS/CEL – G. Guibaud
(Failure Anlysis Expert) Expertise on failure analysis, security, Expertise on failure analysis, security, dependability of electronic components dependability of electronic components +33 (0) 5 6128 1695 gerald.guibaud@cnes.fr www.thalesgroup.com